TRANSCEND-T2D-1 Results: Retatrutide for Type 2 Diabetes (Lancet)

TRANSCEND T2D 1 (Lilly writes it TRANSCEND-T2D-1) is the first Phase 3 trial of retatrutide in type 2 diabetes to report, and the first with a paper, published in The Lancet on June 6, 2026. In 537 adults managing diabetes with diet and exercise alone, average A1C fell 1.69, 1.86 and 1.94 points on 4 mg, 9 mg and 12 mg over 40 weeks, against 0.81 on placebo. Weight fell 11.5%, 13.9% and 15.3%, against 2.6%.

The rest of this page goes through who was in the trial, the results by dose, what else was measured, and what the paper says about side effects, deaths and people who stopped. Where Lilly's March release and the Lancet paper differ, we say which is which.

How was TRANSCEND-T2D-1 designed?

It was a 40-week, double-blind, placebo-controlled Phase 3 trial run at 48 sites in the USA, Mexico and India. Between April 10, 2024 and April 21, 2025, 930 people were screened and 537 were randomized one to one to one to one: 134 to 4 mg, 133 to 9 mg, 136 to 12 mg and 134 to placebo, all by weekly injection under the skin (Lancet; ClinicalTrials.gov record NCT06354660).

Entry required type 2 diabetes that diet and exercise weren't controlling, an A1C between 7.0% and 9.5% and a BMI of at least 23. People had taken no diabetes medicine for at least 90 days, and anyone who had used insulin was out, with the exception of insulin during gestational diabetes (Lilly's March 19, 2026 release). That makes this a monotherapy trial. It's the cleanest test of what retatrutide does by itself, and the least like how most people with diabetes are actually treated.

At the start, people averaged 48.8 years old, 55% were female, diabetes had been present for 2.5 years, A1C was 7.9% and BMI was 35.8, with a mean weight of 96.9 kg (213.6 lb) per Lilly. Doses climbed from 2 mg every 4 weeks: 4 mg after one step, 9 mg via 2, 4 and 6 mg, and 12 mg via 2, 4, 6 and 9 mg. The primary endpoint was the change in A1C at week 40. Percent weight change was a key secondary endpoint.

What were the TRANSCEND T2D 1 results for A1C and weight?

The table gives the Lancet's treatment-regimen estimand, which counts everyone as randomized whether or not they stayed on the drug. The placebo-adjusted A1C differences and their confidence intervals are from the paper. The weight gap against placebo is our subtraction.

Two things stand out. Placebo alone lowered A1C by 0.81 points, so about half of the drop on 4 mg would have happened anyway in this trial. And the A1C curve flattens: going from 9 mg to 12 mg bought 0.08 points, while weight kept falling.

Why do Lilly's numbers look different?

Lilly's release used the efficacy estimand, which asks what happened while people were still taking the drug. On that basis A1C fell 1.7, 2.0 and 1.9 points on 4 mg, 9 mg and 12 mg against 0.8 on placebo, and weight fell 11.5%, 15.5% and 16.8% against 2.5%. That's why you'll see "up to 2.0" in headlines and "1.94" in the paper. Neither is wrong. The efficacy figures are more flattering because they set aside what happened after people stopped, and the treatment-regimen ones are what the Lancet abstract reports. Weight on the efficacy basis was 11.1 kg, 15.1 kg and 16.6 kg lost, against 2.8 kg.

Lilly also said weight loss hadn't plateaued by week 40, so these figures were still moving when the trial ended. The longer 80-week trials are covered on the Phase 3 results page.

How many people reached a normal A1C?

On the efficacy estimand, up to 90% of people on retatrutide got below 7.0%, up to 85% reached 6.5% or below, and up to 46% got below 5.7%, the cutoff under which A1C is outside the diabetes range (Lilly's June 6, 2026 release). "Up to" means the best dose. We didn't find per-dose rates in the sources we opened, so don't read the top figure as typical.

Our read: the under-5.7% number shows what the drug can do at best, not what to expect. These were people about 2.5 years into diabetes with a starting A1C of 7.9%, and someone with a decade of disease and a higher A1C starts from a different place.

What else did the trial measure?

At 40 weeks, up to the best dose, Lilly reported triglycerides down 39.6%, non-HDL cholesterol down 19.8%, systolic blood pressure down 6.4 mmHg and waist circumference down 4.9 inches (12.4 cm). It added that these weren't multiplicity-controlled for 4 mg. The Lancet abstract we could read doesn't list them, so they come from the release. Take them as signals; the kidney, heart and liver questions need longer trials.

For how retatrutide fits into diabetes care overall, including what's still unknown about people on other drugs, see the type 2 diabetes page.

What were the side effects and how many people stopped?

The paper describes the most frequent adverse events as generally mild to moderate gastrointestinal ones that eased over time. Lilly's release gave the rates, and said they clustered during dose escalation.

Nausea climbed with dose; diarrhea and vomiting didn't. Each arm held about 135 people, so small gaps between doses may be noise. Dysesthesia, burning or oversensitive skin, also appears in the other Phase 3 trials Lilly has released, and the skin sensitivity page covers it.

In total, 490 people (91%) finished the 40 weeks on study drug and 504 (94%) completed the study. By subtraction, 47 people stopped study drug early for any reason and 33 left the study altogether. Adverse events explain only a share of that, since stopping for them ran at 2.2% to 5.1% on retatrutide.

Were there deaths or low blood sugar?

Two people died during the study, both in the 4 mg group. The paper says both deaths were unrelated to the study drug. The abstract doesn't give causes or timing, and Lilly's March release didn't mention deaths, so we can't say more than that. With 134 people on 4 mg and none of the higher doses affected, two deaths in one arm of a 40-week trial is a thin base for any conclusion either way.

No severe hypoglycemia was reported. That's expected when nobody is on insulin or a sulfonylurea, so don't carry it over to people who are. Serious adverse event rates weren't in the abstract or in the Lilly releases, and we haven't seen the full paper, so that's a gap.

Call your doctor or get urgent care for severe, persistent abdominal pain, repeated vomiting that keeps you from drinking, or signs of dehydration. Those are class warnings for GLP-1 drugs, not findings specific to this trial.

How does it compare with the other diabetes trials?

Different question, different trial. TRANSCEND-T2D-1 is retatrutide against placebo in early diabetes. The head-to-head against semaglutide's diabetes brand is on the TRANSCEND-T2D-2 page, and TRIUMPH-2 tested people with both obesity and diabetes for 80 weeks (see the TRIUMPH-2 page). Lilly's October 28, 2026 release counts TRANSCEND-T2D-2 as the sixth Phase 3 trial to report.

Putting the numbers next to those for tirzepatide's diabetes trials isn't a fair test, since the trials differed in who they enrolled and how long they ran. The Mounjaro comparison lays out what can and can't be lined up.

Retatrutide isn't approved for anything yet. If you're seeing vials sold online as retatrutide, those aren't Lilly's product and nothing about this trial applies to them; the retatrutide for sale page explains what's actually on offer and what the regulators have said. Lilly last said, on September 29, 2026, that it plans to submit to the FDA in the first quarter of 2027, and the FDA's October 2 decision means the application would go under section 505 rather than as a biologic.